Systems for determining a quota include connectivity devices, a network, a service provider, and a quota management system. The processing load on the service provider and quota management system is driven largely by the number of authorization requests and responses that must be exchanged between the service provider and the quota management system. Therefore sizing the infrastructure for the service provider and quota management system requires an understanding of what the expected peak rate of requests/responses will be (often called TPS or transactions per second). The TPS is determined by the number of connectivity devices times the average number of activities (sometimes called sessions) per device times the average frequency at which each activity or session must reauthorize for a new quota. For instance, in the event that there is a system with 1 million connectivity devices, and each device is engaged in an average of 3 activities, and the quotas being authorized run out or expire in an average of 5 minutes, then the resulting load on the service provider and quota management system will be 3 million quota requests/responses every 5 minutes, which translates to 10,000 TPS.
Many of the service provider and quota management system solutions available were designed originally to work with connectivity devices that primarily supported voice calling. The usage loads (e.g. average calls per day per device) and usage patterns (e.g., peak calls per hour per device) for voice calling across a large set of connectivity devices is quite predictable to within a few percent. In addition, the quota being authorized for voice calls has traditionally been “minutes”, and a voice call would always consume one minute for each minute that the call was connected, so it was obvious how long a given quota would last before a new authorization request was required (e.g. if 10 minutes was authorized, the next request will come after 10 minutes). As a result, all the factors that feed into calculating the peak TPS required from the service provider and quota management system can be controlled or predicted so the necessary infrastructure can be easily allocated to these systems. In addition, a very simple algorithm for determining how much quota to authorize was sufficient to maintain a consistent peak load on these systems (many of these traditional solutions simply had a configured quota quantity that was always applied, e.g. “3 minutes”).
With the advent of smart connectivity devices (e.g. the iPhone), very fast networks (e.g. 4G/LTE), and the myriad of digital services available (e.g. video, HD video, chat, facebook, twitter, email, web browsing, etc.), the usage patterns for connectivity devices have become extremely spiky and unpredictable, and the overall volume of usage has skyrocketed. In addition, the quota authorized for most digital services is a volume of usage, not a time period of usage, so it is no longer obvious how long a quota authorized will last and a new authorization request will be required since it depends entirely on how fast the authorized volume quota is consumed and that depends on many factors such as the power of the connectivity device, the speed of the network, the congestion level of the network, the type of activity, etc. The combination of this extremely high volume but unpredictable usage and the complex relationship between a quota volume and the time it will last before being consumed creates a major threat to the service provider and quota management system since the peak load can no longer be predicted with any accuracy. In the event that the service provider and quota management system infrastructure are not sized sufficiently to handle the peak TPS, service interruptions or system failure will result, causing significant negative business consequences to the service provider.